Is it time to screen all patients with hypertension for primary aldosteronism?

Screening for primary aldosteronism in patients newly diagnosed with hypertension would maximise treatment benefits and minimise end-organ damage

Hypertension is the second leading risk factor for death and disability globally.1 In Australia, 6 million adults are affected2 and, on the basis of the evidence we discuss in this article, a significant proportion may have primary aldosteronism (PA) as a secondary cause of their hypertension. PA, of which a subset has been known as Conn syndrome, is a disease characterised by excessive autonomous aldosterone production.3 Contrary to the traditional teaching that this is both rare and benign, PA is common, specifically treatable, and associated with significant cardiovascular morbidity and mortality.4-6

The high prevalence of PA in patients with severe or resistant hypertension presenting to tertiary centres is well established internationally.7,8 The latest systematic review of over 30 studies conducted in tertiary centres described a prevalence of 16.4% in patients with stage 3 hypertension (systolic blood pressure [BP] > 180 mmHg; diastolic BP > 110 mmHg).9 In contrast, the review identified just nine studies conducted in the primary care setting, where the prevalence ranged from 3.2% to 12.7%, with one small Australian study showing a prevalence of 11.5% among antihypertensive drug trial volunteers.10 More compelling data from a large cohort of 1672 primary care patients in Italy showed a prevalence of 5.9%.11 This is a robust result, given the unselected nature of the patients and the stringent diagnostic criteria. The researchers found that 45% of patients with PA had mild hypertension (systolic BP 140–159 mmHg; diastolic BP 90–99 mmHg), a reflection of their unbiased recruitment. These patients would have remained undiagnosed or been labelled with “essential hypertension” had they not been screened for PA as part of the study.

Undiagnosed PA leads to poor BP control and increased cardiovascular, renal and metabolic morbidity and mortality related specifically to the effect of aldosterone excess mediated by the mineralocorticoid receptor.4 Several research groups have independently shown the detrimental impact of mineralocorticoid receptor activation on cardiac inflammation and fibrosis in animal models.12 Clinically, a study involving 459 patients with PA and 1290 BP-matched controls with essential hypertension, showed an increased prevalence of left ventricular hypertrophy, coronary artery disease, non-fatal myocardial infarction, heart failure and atrial fibrillation in those with PA.13 Another study showed a significant increase in strokes, myocardial infarction and atrial fibrillation in patients with PA compared with those with BP-matched essential hypertension, with odds ratios of 4.2, 6.5 and 12.1 respectively.6

Targeted treatment of PA using surgery or mineralocorticoid receptor antagonists, such as spironolactone and eplerenone, rather than non-specific antihypertensive medications, can reverse the underlying cardiovascular pathology.14 Studies have repeatedly shown that treatment outcomes are better in younger patients with a shorter duration of disease,15-17 highlighting the importance of diagnosing the disease early rather than waiting for the development of severe or resistant hypertension with associated end-organ damage. Furthermore, the accurate diagnosis of a unilateral aldosterone-producing adrenal adenoma permits surgical resection which, when compared with medical therapy, offers a faster onset of BP normalisation, earlier reduction of left ventricular mass, reduced medication use, reduced clinical visits, improved anxiety and depression as well as improved quality of life.18

In PA, the plasma aldosterone level is normal or elevated while renin is suppressed, leading to an elevated aldosterone to renin ratio (ARR) — this ratio is currently recommended as the biochemical screening tool for PA worldwide.19 It is much easier to screen for PA before starting antihypertensive therapy, as many of the commonly used drugs interfere with the ARR.20 If patients are already taking drugs such as β-blockers, calcium channel blockers, angiotensin converting enzyme inhibitors, angiotensin II receptor blockers or diuretics, they often require transition to non-interfering drugs for accurate PA testing.21 The transition process can take weeks and may be difficult in patients with severe hypertension, thereby delaying or even preventing the diagnosis. Active screening of patients newly diagnosed with hypertension is the best way to capture patients with PA early so as to maximise the benefits of treatment and minimise end-organ damage, while avoiding the confounding effect of commonly used drugs. However, the ARR is not a perfect test and may be affected by restrictive or excessive salt intake, oestrogen status, impaired renal function and ageing.4,19 Its diagnostic sensitivity ranges from 64% to 100% and specificity from 87% to 100%, depending on the assay, cut-off threshold, pre-test patient preparation and population characteristics.22 By standardising assays and thresholds for the population to be screened, and accounting for interfering factors when interpreting results, the ARR is a very useful screening test.

A retrospective evaluation of the diagnosis of PA before and after the widespread use of the ARR in five selected tertiary centres around the world revealed a five- to 15-fold increase in the identification of patients with PA.23 The main centre for PA management in Australia, the Endocrine Hypertension Research Centre at Greenslopes Private Hospital and Princess Alexandra Hospital in Brisbane, also found that routine testing of ARR in all patients with hypertension from 1991 led to a dramatic increase in PA diagnoses (Box). This pattern is similarly observed at Monash Health, another Australian centre developing a focused expertise in this field, with PA diagnoses increasing from six in 2011 to 35 in 2016, leading to reduced medication burden with improved BP control for all diagnosed patients (unpublished data).24

While screening patients with hypertension in tertiary centres can increase the detection of PA, most affected patients are not managed in this setting. General practitioners primarily manage hypertension at diagnosis and rarely investigate for PA. A recent survey of 500 GPs from Italy and Germany showed that only 3% of patients with hypertension were screened for PA.25 In Australia, comprehensive BEACH (Bettering the Evaluation and Care of Health) data collected from 15 681 GP participants from April 2000 to March 2016 revealed only 57 cases of PA out of 1 568 100 GP–patient encounters, while an aldosterone measurement was ordered only 66 times over the same period (Christopher Harrison, data custodian of BEACH, personal communication, January 2017). The low rates of diagnosis are also observed in GP clinics throughout Victoria, ranging from 0.1% to 0.45% of patients with hypertension (survey of ten GP clinics in Victoria, unpublished data). Although a younger age of onset of hypertension and associated hypokalaemia may prompt the clinician to consider a diagnosis of PA, both our clinical experience and previous studies have shown that the average age of onset of PA is 40–50 years and hypokalaemia is only present in 9–37% of patients.4 Hence, age and potassium levels are not reliable clinical markers of PA. The evidence from tertiary centres for increased PA diagnosis by screening all patients with hypertension, and the established benefit of screening newly diagnosed, unmedicated patients with hypertension strongly suggest that we should be routinely screening these patients.

The current lack of screening for PA in primary care may be attributed to several factors, including:

• weak or absent guideline recommendations;

• lack of cost-effectiveness analysis and local prevalence data;

• inadequate understanding of PA; and

• complex formal PA diagnosis beyond primary care.

The Endocrine Society recommendation to screen for PA in patients with hypertension who are at higher risk, including those with BP > 150/100 mmHg, controlled BP < 140/90 mmHg taking four or more antihypertensive medications, hypokalaemia, adrenal adenoma, obstructive sleep apnoea or family history of PA,21 is not reflected in other guidelines. The 2016 Australian hypertension guidelines advised clinicians to consider the diagnosis of PA in patients with hypertension, especially those with moderate to severe or treatment-resistant hypertension or hypokalaemia, but did not explicitly recommend active screening in all of the at-risk groups.2 Given the significant burden of hypertension to the Australian health care system, with associated vascular events costing $7.7 billion in 2008–092 and the overwhelming overseas data to support a high prevalence of PA, it would be wise to evaluate the direct costs of introducing ARR to our health care system as a screening test for patients with hypertension and validate it as being cost-effective.

There are only three published cost-effectiveness analyses of screening patients with hypertension for PA. Two of these showed clear cost-effectiveness of screening for PA in patients with resistant hypertension,26,27 while the third calculated a saving of CAD$31 132 for patients with adrenal adenomas to have surgery compared with lifelong medications.28 The cost saving is facilitated by the relatively low cost ($43.70 on the current Australian Medicare Benefits Schedule [www9.health.gov.au/mbs/search.cfm?q=66698&Submit=&sopt=S]) of an ARR. However, there are no economic modelling studies based on the use of the ARR to screen for PA in primary care. Such a study would validate the performance and assess the cost implications of the ARR as a screening test among newly diagnosed patients with hypertension in Australian general practices.

Robust local prevalence data are clearly required if the case for screening is to be made and accepted. There is only one study from Australia where patients were recruited from the community.10 This study was limited by the small sample size (52 participants) and the use of antihypertensive drug trial volunteers rather than actual patients from primary care. A prospective study in which consecutive patients with hypertension recruited by GPs are screened is needed to establish the true prevalence of PA in the Australian primary care setting. The evidence would clearly inform policy making and guideline transformation towards systematic screening for PA.

For clinicians to accept PA as a cause of hypertension worthy of screening, a greater appreciation of the disease, its clinical manifestations and complications is required. Education is also needed on the pathways involved in the formal diagnosis and management of PA, which is often considered complex, time-consuming and invasive. Clinicians who identify an abnormal ARR should be able to refer patients to a streamlined service that offers the patient a definitive diagnosis as well as the most cost-effective treatment. While patients generally require a confirmatory saline suppression test and adrenal vein sampling to delineate the exact cause of their PA, the development of specialist endocrine hypertension centres with a focused expertise would facilitate PA investigation with high efficiency and minimal risk. Even for patients for whom further investigation is not appropriate, an ARR may guide the choice of antihypertensive therapy.

In conclusion, there is growing evidence for a high prevalence of PA with a clear need for early detection so that effective targeted therapy can be instituted. Ongoing studies on the prevalence of PA, together with cost-effectiveness analysis in an Australian primary care setting, should consolidate the evidence for screening all patients newly diagnosed with hypertension. The existing evidence argues for proactive screening for PA before commencement of non-specific antihypertensive therapy. This diagnostic strategy should lead to significant individual and population health and economic impacts as a result of many patients with hypertension being offered the chance of curative or simpler treatment at an early stage of their disease.

Box – Number of patients diagnosed with primary aldosteronism (PA) before and after the introduction of aldosterone to renin ratio (ARR) as a routine screening test in hypertension units at Greenslopes Private Hospital and Princess Alexandra Hospital in Brisbane*

---

* Black = patients with hypokalaemia. White = patients with normal potassium levels.